This application describes a 5 year mentored training program to develop a research career in Child and Adolescent Psychiatry focused on improving phenotypic approaches to defining pediatric bipolar disorder. The principal investigator has completed M.D., Ph.D. and residency training in child and adolescent psychiatry. He will now gain expertise in taxonomic approaches, multi-informant assessment, family and twin designs, and latent variable models. As primary sponsor, Dr. James Hudziak will mentor the Pi's scientific development. Dr. Hudziak is an expert in taxonomic approaches using family study and twin study behavioral genetic designs in concert with latent variable measures to developmental psychopathology. The training program will enlist the co-sponsorship of Dr. Dorret Boomsma, Professor at the Vrije University of Amsterdam and internationally known expert on twin studies and Dr. Thomas Achenbach, Professor at the University, of Vermont. In addition, three internationally-renowned scientists, Drs. Frank Verhulst, Ellen Leibenluft and Michael Neale will provide scientific consultation. Research will focus on diagnostic approaches to children often diagnosed as having pediatric bipolar disorder and related severe dysregulation in affect, behavior and cognition. The training and research will emphasize the use of categorical DSM-based approaches and compare them to quantitative CBCL- based approaches. The proposed studies will combine and contrast DSM and quantitative measures in order to advance the understanding of highly troubled youth. New to this application are proposed research projects allowing for testing of the relations between DSM and quantitative measures, the familiality of DSM and quantitative definitions, the overlap between the approaches, and the construct validity of CBCL-based phenotyping in longitudinal, genetically informative designs. The long term goal of this research is to develop more refined phenotypic approaches for children with severe problems in dysregulation who may or may not meet narrowly defined DSM criteria for pediatric bipolar disorder. If realized, these approaches can then be used in genetic, neuroimaging, and treatment studies aimed at reducing the burden of suffering that children and families with these disorders endure.